What is ChatGPT Go, and how will ads in ChatGPT work?

Yesterday, Friday, January 16, 2026, OpenAI made two announcements that belong together. First, ChatGPT Go, its cheapest paid plan, is now available everywhere ChatGPT is offered, including the United States, where it costs $8 a month. Second, OpenAI plans to start testing ads in ChatGPT in the coming weeks, for adults in the US on the free and Go tiers.

OpenAI presents both as ways to widen access: a cheaper plan for people who want more than the free version, and ads to help pay for free and low-cost use. Fidji Simo, OpenAI's CEO of Applications, wrote the post explaining the ad plans and the rules OpenAI says it will follow.

What ChatGPT Go is

OpenAI launched Go in India in August 2025 as a low-cost subscription. It has since spread to 170 more countries, and OpenAI calls it its fastest-growing plan. With yesterday's change, ChatGPT has three consumer subscriptions worldwide:

Plan	US price per month	What you get
Free	$0	Basic access, with lower limits; ad tests planned in the US
Go	$8	10 times the free tier's messages, file uploads and image creation, using GPT-5.2 Instant; more memory and a bigger context; ad tests planned in the US
Plus	$20	Adds GPT-5.2 Thinking, older models and the Codex coding agent, with higher limits; no ads
Pro	$200	Full access to GPT-5.2 Pro, maximum memory and context, early previews; no ads

Go prices are adjusted to local markets in some countries. Memory here means ChatGPT's ability to remember useful details about you across conversations; context is how much of the current conversation and files it can keep in view at once. Business and Enterprise plans for organizations also remain ad-free.

How the ads will work

No ads have appeared yet. OpenAI describes the first test this way:

Ads will show at the bottom of an answer, only when there is a relevant sponsored product or service connected to your current conversation.
They will be clearly labeled and kept separate from ChatGPT's own answer.
You can see why an ad was shown, dismiss it and say why.
No ads for accounts where the user says, or OpenAI predicts, that they are under 18.
No ads near sensitive or regulated topics such as health, mental health or politics.

OpenAI also lists five principles. The central ones are answer independence, meaning ads don't influence what ChatGPT tells you, and conversation privacy, meaning advertisers don't see your conversations and OpenAI never sells your data to them. You will be able to turn off ad personalization and clear the data used for ads, and OpenAI promises there will always be a way to use ChatGPT without ads, including a paid ad-free tier. It also says it does not optimize for how much time you spend in ChatGPT.

Looking further ahead, OpenAI says conversational ads could become interactive, for example letting you ask questions about a product before deciding to buy.

A worked example

Imagine a free-tier user in the US planning a trip:

Prompt

I'm going to Lisbon for four days in May with my partner. Suggest a
day-by-day plan with neighborhoods to explore, and two mid-priced
places to stay near the old town.

Under the rules OpenAI describes, ChatGPT would answer as usual, with its own suggestions. Below that answer, if a hotel booking service had a relevant sponsored offer, a labeled ad might appear, visibly separate from the itinerary. The ad is not supposed to change which neighborhoods or hotels ChatGPT itself recommended. A question about a health symptom, by contrast, should show no ad at all.

Try it: Write down the last five questions you asked a chatbot. Mark which ones would be eligible for an ad under OpenAI's rules, which fall under the excluded topics, and whether an ad would have helped or distracted you.

Why it matters

Ads could help OpenAI pay for very large spending commitments. CNBC notes that OpenAI signed more than $1.4 trillion in infrastructure deals in 2025, and that Sam Altman said the company was on track for about $20 billion in annualized revenue. Advertising has long funded free services at Google and Meta, and OpenAI is now trying the same model.

It is a notable turn. Altman has spoken in past interviews about his concern that ads could weaken users' trust, though he said in November that he expected OpenAI to try them eventually. Yesterday he wrote on X that plenty of people want heavy AI use without a subscription, and that OpenAI hopes an ad-supported model can serve them.

OpenAI itself acknowledges that people rely on ChatGPT for important and personal tasks. That is why its promises about privacy, sensitive topics and keeping ads separate from answers will be watched closely once the tests begin.

Which ChatGPT plan makes sense for you now?

If you use ChatGPT a few times a week for simple questions, the free tier remains enough, with the caveat that US users may soon see ads under some answers.

If you hit the free limits often, especially for messages, uploads or images, Go at $8 is the new middle step. It gives far more room with GPT-5.2 Instant and better memory, though US Go users will also be part of the ad test.

If you rely on ChatGPT for deeper work, such as long documents, analysis, coding or careful reasoning, Plus at $20 adds GPT-5.2 Thinking and Codex and has no ads. Pro at $200 is for heavy professional users who need the most capable model and the highest limits.

A practical approach is to start one level lower than you think you need, note where you run into limits over a couple of weeks, and upgrade only if they get in your way.